This issue was fixed in Oracle Solaris 11.3.2.4.
If you reboot the root domain with I/O Resiliency (IOR) enabled (that is, logical domains failure policy set to ignore) the I/O domain might fail to initiate an ixbgevf network interface.
For example, before rebooting non primary root domain (NPRD), type:
# ipmpstat -i INTERFACE ACTIVE GROUP FLAGS LINK PROBE STATE net1 yes ipmp0 --mbM-- up disabled ok net7 yes ipmp0 ------- up disabled ok # dladm show-phys LINK MEDIA STATE SPEED DUPLEX DEVICE net0 Ethernet up 0 unknown vnet0 net7 Ethernet up 10000 full ixgbevf7 net1 Ethernet up 10000 full ixgbevf4
After rebooting the NPRD, type:
# ipmpstat -i INTERFACE ACTIVE GROUP FLAGS LINK PROBE STATE net1 no ipmp0 ------- down disabled failed net7 yes ipmp0 --mbM-- up disabled ok # dladm show-phys LINK MEDIA STATE SPEED DUPLEX DEVICE net0 Ethernet up 0 unknown vnet0 net7 Ethernet up 10000 full ixgbevf7 net1 Ethernet down 10000 full ixgbevf4
Recovery: Perform these steps.
Unplumb and plumb the failed interface.
# ipadm delete-ip net1 # ipadm create-ip net1
Add the device into the ipmp group again.
# ipadm add-ipmp -i net1 ipmp0